Incorporated: 1984 as Amtec Securities Corp.
NAIC: 334413 Semiconductor and Related Device Manufacturing
SIC: 3674 Semiconductors & Related Devices
Ramtron International Corporation is a NASDAQ-listed semiconductor company based in Colorado Springs, Colorado. Ramtron designs, develops, and markets ferroelectric random-access memory (FRAM) products. FRAM combines the best features of the two primary semiconductor memories, random-access memory (RAM) and read-only memory (ROM). Although RAM is high performing, easy to use, capable of being rewritten countless times, and inexpensive, it loses its contents when power is removed. ROM, in contrast, is more stable but expensive, slow to write, easy to wear out, and demands a great deal of power. FRAM tries to bridge this gap by taking RAM and adding a layer of ferroelectric materials, ceramic crystals, which become polarized if charged and remain in that state once the electric field is removed. This trait, as a result, allows for the creation of nonvolatile memory cells. The basic idea has been around for many years, and has tantalized the semiconductor industry for just as long, but Ramtron and others have yet to make it a truly commercial technology.
Ramtron products are manufactured by a foundry in Japan and sold to original equipment manufacturers for use in such products as meters, from utility to taxi; computing and information products, such as servers and laser printers and copiers; communications devices, such as portable Global Positioning System (GPS) and cell base stations; consumer products, including plasma and LCD televisions and set-top boxes; industrial, scientific, and medical instruments; and automotive components subjected to high temperatures that accelerate data loss.
Although still a small company after being in business for 20 years, Ramtron does not lack major supporters. It has strategic alliances with Fujitsu Limited, Hitachi Ltd., Infineon Technologies AG, Rohm and Haas Company, Texas Instruments Inc., and Toshiba Corporation. Ramtron maintains North American sales offices in Montreal, Dallas, and San Diego, and covers Europe and Asia from Bracknell, United Kingdom; Hong Kong; and Kanagawa, Japan. The company also sells its products through an extensive network of distributors around the world, with the exception of the Middle East.
Ferroelectrics: 17th-Century Roots
The first known ferroelectric material, sodium potassium tartrate tetrahydrate, better known as Rochelle salt, was first separated by an apothecary in La Rochelle, France, in 1655. It was used as a mild purgative for the next two centuries before scientists discovered its pyroelectric traits, the ability to become polarized when charged by a nearby electric field. During World War II the pace of research accelerated with the discovery of barium titanate. Many more ferroelectrics were then identified over the next decade and the capacitor and transducer applications of these materials were researched. In the 1950s the researchers at IBM, Bell Labs, and Westinghouse began investigating ways to apply ceramic crystals to computers.
Ramtron grew out of the work of George Rohrer, who in 1968 while a graduate student at Michigan Technological University developed the first polycrystalline thin film ferroelectric memory using potassium nitrate. Raising funds from Detroit investors, he formed Technovation Corporation to commercialize his patented technology. The task proved difficult; Rohrer soon ran out of money, but was able to raise more funds from a large group of investors from Sault Sainte Marie, Michigan. In time Rohrer was again broke and forced to put a halt to his research. He was not alone in his failure to commercialize FRAM, however. By the early 1970s IBM was ready to declare FRAM a pointless pursuit. The ceramic crystals did not react well with silicon, the building block of memory chips, and the FRAM prototype was unable to hold a charge for very long, suffering from "fatigue."
Formation of Ramtron Corporation: 1984
Nevertheless, FRAM continued to hold out great promise, and in 1983 a group of Australian venture capitalists approached Rohrer about buying out Technovation and giving his research another try. Rohrer recruited Larry D. McMillan, a manager at Honeywell who brought in a friend, Dr. Carlos Araujo, a professor at the University of Colorado in Colorado Springs (UCCS). In 1984 they and their Australian backers formed Ramtron Corporation, the "Ram" portion of the name drawn from the initials of Rohrer, Araujo, and McMillan.
Ramtron started out as a research and development company, setting up shop in a UCCS laboratory in May 1984 to develop ferroelectric materials compatible with computer semiconductors. The investors installed a group of managers who were soon at odds with Rohrer, Araujo, and McMillan over the ferroelectric material to use, the newcomers electing to pursue lead zirconate titanate (PZT) rather than potassium nitrate. The researchers did manage to develop a new design to eliminate the incompatibility with silicon, essentially devising a FRAM sandwich with microscopic electrodes to create a buffer between the ferroelectric material and the silicon base.
However, the founders were not happy with their supporting roles. McMillan and Araujo left the company in 1986 to form Symetric Corporation, which focused on FRAM's fatigue problem, giving themselves five years to find a solution. Shortly before their self-imposed deadline, they found a silicon substitute, strontium bismuth tantalite, which was able to perform for a decade without suffering from fatigue. Rohrer, in the meantime, could never accept the change in materials after devoting so much of his life to potassium nitrate, and he sold the Technovation's intellectual property and left Ramtron in 1987. He moved to the Upper Peninsula of Michigan where he lived until his death at the age of 62 in 1998.
Amtec Securities Corp., a subsidiary of the Australian company Newtech Development Corporation Limited formed in 1984 by Ross Lindon-James and Brian Harcourt, acquired a half-interest in Ramtron in 1986, then acquired the rest of the company two years later and changed its name to Ramtron International Corp., while parent company Newtech took the name Ramtron Australia Ltd. In 1992 Ramtron Australia was dissolved and shares of Ramtron International were distributed to its shareholders.
Ramtron's research efforts continued in the meantime, aided in large measure by the 1989 infusion of $12 million by the pension fund of the National Electrical Contractors Association (the National Electrical Benefit Fund), which became the company's largest shareholder. Another important investor in the early 1990s was Colorado businessman Oren L. Benton, who owned a 40 percent stake after lending money to the company in 1993, and became chairman and chief executive officer in June 1994.
For several years Ramtron appeared to be on the verge of introducing a breakthrough product to the market but nothing came to fruition. Nevertheless, Ramtron was able to forge alliances with some important electronics companies. In May 1992 it signed an agreement with Hitachi to determine the feasibility of high-density FRAM products. Following a successful feasibility study, the two parties expanded the alliance in April 1994 by signing a five-year development agreement. A year earlier Ramtron also established an alliance with semiconductor company Rohm Company, Ltd., to develop low-density FRAM products, and this too led to an expanded licensing and manufacturing agreement. A development and manufacturing agreement with Toshiba was then struck in July 1995. The following year Samsung Electronics Co., Ltd., signed a license agreement to manufacture and sell FRAM products, and in that same year Fujitsu Ltd. became a fourth licensee of the core technology. As a result, Ramtron moved beyond a mere research and development company and began to post income from product sales, royalties, and license agreements. In 1992 revenues totaled less than $1.5 million, but two years later that amount increased to more than $20 million, and in 1995 revenues improved to $28.9 million and the company was able to cut its annual loss from $20 million to $2.5 million.
Ramtron experienced a setback in 1995 through no fault of its own. Chairman and CEO Benton had made his fortune as a uranium trader, but another company he headed, Uranium Resources Inc., encountered problems, saddling him with about $500 million in personal debt. In February 1995 Benton filed for Chapter 11 bankruptcy protection. Benton's bankruptcy had an impact on Ramtron's efforts to restructure its debt in 1995. Because Benton's assets were involved, the company had to receive approval from bankruptcy court in order to trade stock for about $24.3 million in debt.
Shortly after declaring bankruptcy, Benton resigned as Ramtron's CEO and chairman, and as a director. He was replaced on an interim basis by his predecessor, George J. Stathakis, before a permanent replacement was found in L. David Sikes, who had resigned as Ramtron's president shortly before Benton's problems emerged. Sikes, who held a degree in electrical engineering from the Massachusetts Institute of Technology, had spent 18 years with Motorola, Inc., as well as stints at Eastman Kodak, National Semiconductor Corporation, ASM Epitaxy (a semiconductor company), and ASM America, a semiconductor equipment company where he served as president and CEO. He had first joined Ramtron in July 1992, becoming president and chief operating officer. As Ramtron's CEO, it was Sikes who decided that it would be wiser if Ramtron licensed its technology rather than manufacturing chips itself, and the little manufacturing that was done at the Colorado Springs facility was phased out.
Establishing Enhanced Memory Systems: 1995
In addition to FRAM products, Ramtron also developed Enhanced DRAM (EDRAM) products in the 1990s. To operate this business, Ramtron formed a subsidiary, Enhanced Memory Systems, Inc. (EMS), in 1995. The company's EDRAM products also found some important manufacturing partners in IBM and Nippon Steel. For a time in the mid-1990s Ramtron's focus actually shifted from FRAM to EDRAM products. The sale of EDRAM products increased from $10 million in 1995 to $16 million in 1996, just more than half of the company's total revenues of $31.4 million. Ramtron continued to lose money on the year, about $5.7 million in 1996, but it appeared that the company was on the verge of finally turning the corner. Revenues tailed off in 1997, however, falling to $20.5 million and leading to another loss of $8.8 million.
In 1998 the worldwide semiconductor industry experienced a difficult year, and Ramtron's inability to commercialize some of its FRAM products resulted in an even more disappointing year for the company. Ramtron was at least able during the year to shut down manufacturing at its high-cost Colorado Springs facility, which was not dedicated to research and development, while manufacturing was shifted to less expensive external foundries. However, that provided cold comfort when scanning the balance sheet, which revealed a dip in revenues to $18.5 million and a net loss of more than $19.1 million.
Ramtron closed the 1990s with a mixed-bag performance. The share price plummeted so far that the company engineered a five-to-one reverse stock split in order to sustain the share price high enough to keep the attention of Wall Street. Nevertheless, Ramtron enjoyed success on a number of fronts in 2000 when revenues rebounded to $24.9 million and the net loss was reduced to $2 million. The company also set itself up for further growth by introducing eight new products during the year and establishing five alliances for both FRAM and EDRAM technologies.
Positive news continued in 2000 when German chip manufacturer Infineon Technologies acquired a 20 percent stake in EMS and agreed to provide as much as $200 million per year in DRAM (dynamic RAM) foundry production. EMS was also strengthened by the acquisition of Mushkin Inc., a Denver Internet retailer of high-performance memory modules, whose addition was expected to increase the distribution of the unit's EDRAM products. Economic conditions were not the best, however, resulting in a modest increase in revenues to $26.1 million and another loss of $14.5 million.
A new CEO took over in 2001, as Sikes turned over the reins to William Staunton III, who boasted almost three decades of experience in the high-tech and semiconductor industries. Difficult times in the high-technology sector continued in 2001, though, when demand for semiconductors in general fell, and Ramtron's sales slipped to $22.9 million and its net loss ballooned to $33.2 million. Nevertheless, the company continued to make strides in transforming FRAM from a niche technology into a mainstream one. In 2001 Texas Instruments became the eighth semiconductor company to sign licensing agreements and NEC became the ninth.
Ramtron enjoyed a banner year in 2002, posting revenues of $50.5 million and narrowing its net loss to under $2 million. However, it was unable to build on the momentum the following year, when sales slipped to $42.4 million and the company lost another $9.5 million. The year also brought the retirement of Sikes as chairman, replaced by William G. Howard, Jr., a veteran of the semiconductor and microelectronics industries since the 1960s.
Ramtron took steps in 2004 to sell EMS and refocus its efforts on its original FRAM business, which experienced an excellent 2004. As a result, total revenues soared to $57.8 million and the company was able to finally turn a profit, netting $3.6 million for the year. To further grow the business, Ramtron acquired Montreal-based Goal Semiconductor for $7.4 million in cash and stock. Goal brought analog and mixed-signal chip technology that Ramtron hoped to combine with its FRAM chips to create new products for such markets as power meters, laser printers, and handheld devices. While sales from continuing operations fell from $39.5 million in 2004 to $34.4 million in 2005, Ramtron appeared to be on the right track in finding niche opportunities for its FRAM products. Mushkin was also sold in 2005. After losing $6.5 million in 2005, the company returned to profitability in 2006, netting $500,000 on revenues that increased to $40.5 million.
Principal Subsidiaries
Ramtron K.K.
Principal Competitors
Atmel Corporation; Samsung Electronics Co., Ltd.; STMicroelectronics NV.
Further Reading
Bainbridge, Kim, "Ramtron Deal Adds Products," Colorado Springs (Colo.) Gazette, August 31, 2005.
Cross, L. E., and R. E. Newham, "History of Ferroelectrics," Ceramics and Civilization, Vol. 3, High Technology Ceramics--Past Present, and Future, Westerville, Ohio: American Ceramic Society, Inc., 1987.
Hirschman, Dave, "Investors' Faith in Ramtron's Technology Begins to Pay Off," Colorado Springs (Colo.) Gazette, March 7, 2000, p. BUS1.
Keeton, Laura E., "Ramtron's Benton Quits As Chairman, CEO and Director," Wall Street Journal, February 27, 1995, p. B4.
McMillan, Larry D. "Inventor, Innovator, Entrepreneur and Corporate President: Industrialization of the FeRAM," unpublished dissertation, 2005.
Walsh, Chris, "No Flash in the Pan," Colorado Springs (Colo.) Gazette, August 6, 2000, p. BUS1.
------, "Ramtron Chairman Retires After 10 Years," Colorado Springs (Colo.) Gazette, January 29, 2003, p. BUS 1.
Weilman, Wayne, "Ramtron Weathers Storms, Quake," Colorado Springs Gazette, February 16, 2007.
Weinberg, Neil, "A Computer in Every Shirt Collar?" Forbes, March 8, 1999.
— Ed Dinger
